Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

100x defect tolerance: How we solved the yield problem

cerebras.ai

Post reply on HN
Print view View original

131–140 of 186 posts

Re: 100x defect tolerance: How we solved the yield problem

#131

Post by immibis »

mschuster91 wrote: post #81
__Joker wrote: post #75

Earlier quoted context omitted.

"While I continue to believe that many people are going to collectively lose trillions of dollars ultimately pursuing "AI" at this stage" Can you please explain more why you think so ? Thank you.

It's a hype cycle with many of the hypers and deciders having zero idea about what AI actually is and how it works. ChatGPT, while amazing, is at its core a token predictor, it cannot ever get to an AGI level that you'd assume to be competitive to a human, even most animals. And just as every other hype cycle, this one will crash down hard. The crypto crashes were bad enough but at least gamers got some very cheap GP…

Why do you think that an AGI can't be a token predictor?

Re: 100x defect tolerance: How we solved the yield problem

#132

Post by mschuster91 »

immibis wrote: post #131
mschuster91 wrote: post #81

Earlier quoted context omitted.

It's a hype cycle with many of the hypers and deciders having zero idea about what AI actually is and how it works. ChatGPT, while amazing, is at its core a token predictor, it cannot ever get to an AGI level that you'd assume to be competitive to a human, even most animals. And just as every other hype cycle, this one will crash down hard. The crypto crashes were bad enough but at least gamers got some very cheap GP…

Why do you think that an AGI can't be a token predictor?

Because an LLM _by definition_ cannot even do basic maths (well, except if you're OpenAI and cheat your way around it by detecting if the user asks a simple math question).

I'd expect an actually "general" intelligence Thing to be able to be as versatile in intellectual tasks as a human is - and LLMs are reasonably decent at repetition, but cannot infer something completely new from the data it has.

Re: 100x defect tolerance: How we solved the yield problem

#133

Post by pezezin »

pants2 wrote: post #96
twic wrote: post #87

Earlier quoted context omitted.

On a tangent: has anyone built an active cooling system which operates in a partial vacuum? At half atmospheric pressure, water boils at around 80 C, which i believe is roughly the operating temperature for a hard-working chip. You could pump water onto the chip, have it vapourise, taking away all that heat, then take the vapour away and condense it at the fan end. This is how heat pipes work, i believe, but heat pip…

No need to bother with a partial vacuum when ethanol boils at around 80 C as well and doesn't destroy electronics. I'm not aware of any active cooling systems utilizing this though.

May I introduce you to the glorious vodka cooled PC? https://www.youtube.com/watch?v=IYTJfLyo_vE

Re: 100x defect tolerance: How we solved the yield problem

#135

Post by versteegen »

mschuster91 wrote: post #132
immibis wrote: post #131

Earlier quoted context omitted.

Why do you think that an AGI can't be a token predictor?

Because an LLM _by definition_ cannot even do basic maths (well, except if you're OpenAI and cheat your way around it by detecting if the user asks a simple math question). I'd expect an actually "general" intelligence Thing to be able to be as versatile in intellectual tasks as a human is - and LLMs are reasonably decent at repetition, but cannot infer something completely new from the data it has.

Define "by definition".

Because this statement really makes no sense. Transformers are perfectly capable (and capable of perfectly) learning mathematical functions, given the necessary working-out space, e.g. for long division or for algebraic manipulation. And they can learn to generalise from their training data very well (although very data-inefficiently). That's their entire strength!

Re: 100x defect tolerance: How we solved the yield problem

#136

Post by kragen »

dylan604 wrote: post #37
bigmattystyles wrote: post #30

When I was a kid, I used to get intel keychains with a die in acrylic - good job to whoever thought of that to sell the fully defective chips.

wow, fancy with the acrylic. lots of places just place a chip (I'm more familiar with RAM sticks) on a keychain and call it a day.

Those aren't just a chip; they're an epoxy package with a leadframe and a chip inside it. To put just a chip on a keychain, you'd have to drill a hole through it, which is difficult because silicon is so brittle—almost like drilling a hole in glass. Then, when someone put it onto a keyring, the keyring would form a lever that applies a massive force to the edge of the brittle hole, shattering the brittle silicon. Potting the chip in acrylic resin is a much cheaper solution that works better.

Re: 100x defect tolerance: How we solved the yield problem

#137

Post by kragen »

ChuckMcM wrote: post #102
bigdict wrote: post #100

Earlier quoted context omitted.

Why are you assuming bad faith?

What gave you the impression I was assuming bad faith? It's off topic to the discussion (which is fine) but can be annoying in the middle of an HN thread.

You said, "I would guess you're not asking a serious question here," which is to say, you were guessing that the question was asked in bad faith. Or, at any rate, you would, if for some reason the question came up, for example in deciding how to answer it. Which is what you were doing. That is to say, you did guess that it was asked in bad faith. Given the minimal amount of evidence available (12 words and a nickname "__Joker") I think it's reasonable to describe that guess as an assumption. Ergo, you were assuming bad faith.

Re: 100x defect tolerance: How we solved the yield problem

#138

Post by sharpesttool »

ryao wrote: post #54
sroussey wrote: post #15

Earlier quoted context omitted.

I wonder if you could… just not cut the wafer at all??

I suspect this would cause alignment issues since you could literally rotate it into the wrong position when doing soldering. That said, perhaps they could get away with cutting less and using more.

You just need a sharpie to mark the top.

Re: 100x defect tolerance: How we solved the yield problem

#139

Post by dogcomplex »

mschuster91 wrote: post #132
immibis wrote: post #131

Earlier quoted context omitted.

Why do you think that an AGI can't be a token predictor?

Because an LLM _by definition_ cannot even do basic maths (well, except if you're OpenAI and cheat your way around it by detecting if the user asks a simple math question). I'd expect an actually "general" intelligence Thing to be able to be as versatile in intellectual tasks as a human is - and LLMs are reasonably decent at repetition, but cannot infer something completely new from the data it has.

Yet they can get silver medal PhD level competition math scores.

Perhaps your "definition" should be simply that LLMs have temporarily seen limitations in their ability to natively do math unassisted by an external memory, but are exceptionally good at very advanced math when they can compensate for their lossy short-term attention memory...

Post reply on HN